Our previous work has shown that the invasive and migratory potential of human prostate carcinoma cells is determined in part by the surface expression of the a6 integrin. On normal prostate cells the a6b4 and the a6b1 integrins are present. On prostate tumor cell surfaces within the human tissue samples, the a6b1 integrin dominates. Our current working hypothesis is that the continued expression of the a6b1 integrin on prostate tumor cell surfaces allows for invasion of tumor cells along laminin coated structures such as vessels and nerves. The invasion of the tumor cells occurs along these structures through the action of a functional a6b1 integrin laminin receptor. Strategies for interruption invasion includes expression of dominant negative mutations of b4 and inhibiting the adhesion properties of the a6b1 integrin. The specific aims are: 1. Determine if transfection of the b4 integrin alters the invasion phenotype of the DU145 H prostate cancer cells using the SCID mouse model system. Different constructs of b4 will be used to test whether any changes in phenotype is dependent upon the signaling domain and if the b4 unit without the cytoplasmic domain will decrease the expression of a6b1 and block the invasion phenotype. 2. Determine if the a6t forms of a6, which is present on tumor cells, is an active receptor for invasion and/or migration. These experiments include determining which portion of the a6 molecule is missing and transfection of the altered a6 form into a a6 minutes prostate cancer cells. The participation of the a6t in the invasion process will be tested using the SCID mouse model system. 3. Determine if binding of the cell surface a6 will black the invasive phenotype. Synthetic small molecule ligands and peptides obtained from a combinatorial library will be screened by their ability to bind to a6 on the cell surface and their ability to inhibit a6 dependent adhesion to laminin. The ability of the molecules to inhibit invasion will be tested using a SCID mouse model system.